Herbal medicines have been used as a treatment option for rheumatic disease (RD), but they often produce liver enzyme abnormality. This study examines the incidence of herb-induced liver injury (HILI) and the relationship between risk factors and liver enzyme abnormality (LEA) in inpatients with RD. HILI was analyzed using the Roussel Uclaf causality assessment method liver injury criteria and causality assessment. Multivariable analysis was performed to assess the relationship between patient characteristics and LEA in RD. The features of LEA were also examined in each RD. Among 352 patients included in this study, 105 patients showed LEA on admission, of which 6 had fulfilled the Roussel Uclaf causality assessment method criteria. The incidence risks of LEA and HILI were 12.55% and 0.58%, respectively. Multivariable analysis showed that LEA on admission and occasional use of alcohol could be risk factors for LEA on follow-up. In an additional analysis with each RD, all rheumatoid arthritis patients with LEA were taking nonsteroidal anti-inflammatory drugs, steroids, and disease-modifying antirheumatic drugs, and 4 out of 5 gout patients with LEA were taking steroids. The use of herbal medicine in RD is relatively safe. However, regular monitoring of liver enzyme tests and examination of alcohol consumption are required.